292
now become available for this application [ 11 ]. More recently,
concern has been expressed that removal of carrier proteins
like albumin will also remove less abundant proteins of interest
as a result of a sponge effect caused by the ability of albumin
and other high abundance carrier proteins to bind peptides or
protein fragments and significant differences are found to
occur between different depletion methods used [ 12 ].
Investigation of a commercial albumin depletion kit showed
that low abundance proteins were significantly reduced by
albumin depletion [ 13 ].
- Selected samples of excised gel bands can be sent to a specialist
protein fingerprinting facility for further analysis [ 14 ]. In one
such procedure, proteins are in-gel digested with trypsin and
the digest analyzed by nLC-MS-MS using a 4000 QTRAP
(Applied Biosystems). The observed peptides can be analyzed
using the MASCOT online search database [ 15 ] against the
appropriate species-specific databases (e.g., NCBInr and IPI).
Peptides that have ion scores above the significance threshold
are reported and then these are grouped into their respective
protein identifications, with Mowse scores indicating the prob-
ability of correct identification.
Proteins that were initially identified with two or less peptides
and Mowse scores close to the probability threshold can
be subsequently confirmed and validated by an ion inclu-
sion method. In this mode, theoretical lists of tryptic
digest peptides of known m/z values are generated from
the known protein sequence. They are used as a target list
during repeat Nlc-MS-MS analysis on a LTQ Orbitrap XL
Ion Trap (Thermo Fisher Scientific, Loughborough, UK).
An alternative method suitable for abundant low molecular
weight peaks is to test using a SELDI chip processor
linked to MALDI MS/MS in linear and reflectron mode.
This method allows direct identification of the peak iden-
tified on the SELDI chip.
Table 3
Statistical techniques applied by SELDI-TOF MS analyzer biomarker
wizard software
Number of sample groups Parametric test Non- parametric test
Two Two-tailed t-test Mann-Whitney t-test
Two paired sample groups Paired t-test Wilcoxon t-test
Three or more One-way ANOVA Kruskal-Wallis h-test
Peter Kelly